OpenAI has quietly released text-davinci-003 , which is a brand new entry in the Artificial intelligence The GPT-3 language model family, it claims it can handle more complex prompts to produce longer outputs.
However, as I mentioned before Ars Technica (Opens in a new tab)enterprising users using stadium (Opens in a new tab)However, the free demo of GPT-3 soon found that the new model was more adept at producing poetry and lyrics.
Hacker news commenters have found (Opens in a new tab) That he could write poems about Einstein’s theory of relativity, and then rewrite them in the style of the romantic poet John Keats. As Professor Ethan Mullick, a professor at the Wharton School of Business at the University of Pennsylvania, speaks lyrically about the possibilities in Tweets series (Opens in a new tab).
GPT-3 art generation, and the future
The improvements to GPT-3’s understanding of rhyme and meter are likely due to its filling in with more reference material. The Github repository for GPT-3 (Opens in a new tab) It acknowledges that it draws its massive collection from thousands of data sets.
enthusiasts have them pointed out (Opens in a new tab) Previous iterations of GPT-3 had some quick realization of rhyme schemes, but this latest step forward is a sign that this Amnesty International writer Now he has a complex understanding of meter, and can finally compose artworks of his own.
While all of these developments are exciting, they also raise the question of how human artists, writers, and journalists can exist alongside a technology that is becoming less and less of a “bleeding edge” with each passing day.
The fear for some is that AI, which is able to write and rewrite faster than any human, will take away the paid work from them. Creating and manipulating text using artificial intelligence is nothing new, with tools like Language is a virus (Opens in a new tab) And GPT-3 powered InferKit (Opens in a new tab) After I offered some versions of these posts for some time.
While it’s true that AI text generators (and art generators like DALL E) take a lot of work out of creativity, claims still have to be formulated by humans. And when it comes to being able to replenish certain parts of output as someone dictates (“inpainting (Opens in a new tab)’, which both DALL E and alternative stable diffusion can do), this is also a human-driven process.
So instead of seeing the latest batch of truly competent AI generation systems as a threat to human creativity, we can think about how we can accommodate and collaborate with it.
This could be as a means of inspiration, making artistic processes accessible to more people, or creating entirely new human/hybrid processes.
And if you still feel threatened, consider this: If the content you produce is repeatable by an AI system… wouldn’t you rather make something else? The new and improved GPT-3 (or GPT-4, which is rumored to be on the horizon) could give you more time to do this.